Host-Microbiota Interaction in Chronic Inflammation and CRC
Host-Microbiota Interaction in Chronic Inflammation and CRC
The human intestinal tract is home to a complex ecosystem of commensal bacteria that live in a mutually beneficial state with the host. The number and composition of normal flora vary in different regions of the GI tract. A relatively low number and few species live in the stomach and upper small intestine due to the specific composition of luminal ingredients and the propulsive motion of the region. However, the distal part of the small intestine and colon is habitat to a diverse and densely populated microbiota. Phylogenetic analysis reveals a similar bacterial content in the distal ileum, ascending colon and rectum. The study of intestinal microbiota has gained considerable attention in recent years. The normal flora are composed of bacteria, archaea and fungi. To date, 500 different species of bacteria have been detected in the intestinal tract. The highest microbiota content is found in the distal ileum and colon where the bacterial concentration reaches 10–10 cells/ml of the luminal content. Under physiologic conditions, the host and microbiota retain a mutually beneficial symbiosis. However, any dysbiosis due to either microbiota composition alteration or host genetic susceptibility could initiate an unwarranted immune response.
The epithelial monolayer is covered with a mucus layer, which, together with the epithelium, forms the physical barrier against the luminal content. The mucus layer is 100–250 µm thick in the stomach and proximal small intestine, whereas it is 450 and 700 µm thick in the ileum and colon, respectively. The mucus layer is mainly composed of mucin glycoproteins and trefoil peptides. MUC2, MUC3, MUC12, MUC13 and MUC17 have been identified in the intestinal tract, with MUC2 being the predominant component of the mucus layer. While some mucins are secreted, some are membrane bound and reside on the apical surface of the epithelium. The mucus layer is composed of two distinct layers, both of which are mainly composed of MUC2. Mucus layers of the small intestine and colon are different in a number of ways. The small intestine does not have two distinct layers. In addition, the mucus layer is not firmly attached to the epithelium and can be relatively easily removed. By contrast, the mucus layer in the colon forms two distinct layers; the inner of which is firmly attached to the epithelium. The outer layer is more voluminous and is loosely attached to the inner layer. Under normal conditions, the inner layer is impermeable to bacteria.
In the noninflamed colon mucosa, bacteria are normally present within the crypts in the proximal colon. Moreover, the microbial composition of the crypts is found to be distinct from that of the lumen. However, the effect of this crypt-specific core microbiota on tissue homeostasis is currently unknown. In addition, despite knowledge on the fecal microbiota composition alteration in CRC and IBD patients compared with healthy controls, it is not known whether the crypt-specific core microbiota is also altered. However, a significant difference in microbiota composition has been observed between stool and mucosal biopsy samples in IBD. Spatial organization of bacteria is found to be altered in dextran sulfate sodium (DSS)-induced and IL-10 spontaneous colitis in mice and in IBD patients, as a consequence of which higher numbers of bacteria are found in the crypts. This change brings the bacteria into the close vicinity of stem cells at the base of crypts. Occasionally, intracellular bacteria are also observed in single epithelial cells; the pathologic relevance of this to intestinal homeostasis is unknown.
A number of studies have investigated the effect of intestinal injury by nonlethal intestinal bacteria on gut homeostasis and ISC in Drosophila. A net increase in the quantity of undifferentiated cells (ISCs and enteroblasts) is observed after intestinal injury. However, the number of ISCs is only mildly increased. In other words, the injury results in augmented stem cell division without significant expansion of the stem cell compartment. In addition, the stem cell features – that is, self-renewal and differentiation capacity – are not affected. Interestingly, TLR signaling is not altered at the transcriptomic level in this study. Buchon et al. suggest that ISCs are responsive to the damage signals rather than bacteria. For example, the effect of nonlethal intestinal infection on stem cell division is found to be mediated by reactive oxygen species and unpaired3, a component of the JAK/STAT pathway, which is produced by the damaged cells. The direct effect of microbiota on ISCs has been studied in axenic Drosophila. Axenic flies demonstrate a lower level of stem cell division and epithelium turnover, which are suggestive of the effect of microbiota on the ISC physiology. Notably, unpaired3 is completely absent in axenic flies while it is expressed in conventional counterparts.
Despite the recently elucidated significant difference in the genomic response to systemic inflammation in humans and mice, murine models are widely used to mimic human conditions. When wild-type mice are colonized with Salmonella typhimurium, intestinal epithelial proliferation is increased, which is more predominant at the base of the crypts. In addition, the intestinal expression of Wnt and Notch pathway components are affected.S. typhimurium has been shown to activate the Wnt signaling pathway both in vitro and in vivo. Short-term in vitro incubation of rat small intestinal epithelial IEC-18 with S. typhimurium results in a one- to sixfold increase in the gene expression level of Wnt2, Wnt2b, Wnt3a, Wnt7b and Wnt10b. The same experiment in HCT116 cells results in a two- to tenfold increase in the gene expression level of Fzd3, Fzd4, Fzd6, Fzd9 and Fzd10. Wnt2 and Fzd9 showed the highest alteration at the gene expression level. Interestingly, when human epithelial cells were manipulated to overexpress Wnt2, a lower level of NF-κB activity and downstream cytokine secretion was observed after S. typhimurium incubation. Intestinal Wnt2 expression at the protein level was also found to be increased in germfree mice colonized with S. typhimurium compared with control. Axin1 is a negative regulator of Wnt pathway. In vitro incubation of HCT116, HT29C19A and CaCo2BBE human intestinal cell lines with S. typhimurium significantly decreased Axin1 at the protein level while the gene expression level remained unchanged. This alteration was specific to S. typhimurium and was not reproduced by a proinflammatory stimulus such as TNF-α or nonpathogenic bacteria including Escherichia coli F18 and Lactobacillus rhamnosus GG. Similar reduction in Axin1 protein level was also observed in vivo. The effect of S. typhimurium on the intestinal epithelium was found to be via post-transcriptional modification of Axin1. Larsson et al. have conducted systematic intestinal gene expression analysis of germfree versus colonized mice. We queried the publicly available data for Wnt pathway ligands and receptors (Figure 1>). The genes that were significantly altered (p < 0.05) in the colon of germfree versus colonized mice were noted. Wif1, Wnt5a, Wnt8b, Fzd3, Fzd8, Sfrp4, Sfrp2 and Dkk3 were significantly altered in germfree mice compared with conventional counterparts. The protein–protein interaction map of these proteins with other members of the Wnt pathway is illustrated in Figure 2. The net effect of microbiota on the Wnt pathway cannot be inferred from the available data. Nonetheless, the data clearly show a potential regulatory mechanism of the Wnt pathway by the microbiota and plausibly by any alteration in the microbiota composition.
(Enlarge Image)
Figure 1.
Canonical Wnt signaling pathway. Wnt ligands undergo lipid modification in the endoplasmic reticulum of the Wnt-producing cells. PORCN is expressed in Wnt-secreting cells and is required for Wnt maturation and secretion. On the target cell, Wnt ligand binds to the Frizzled/LRP5,6 heterodimer. GSK and CK1 are kinases that phosphorylate components of the Wnt signaling cascade. sFRPs and WIF bind to Wnt ligands and DKK binds to LRP5/6, and thus inhibits pathway activation. Norrin and R-spondin are not components of the canonical Wnt pathway. However, they function via FZD/LRP and augment Wnt signaling. LGR is a R-spondin receptor and physically interacts with the FZD/LRP complex. In the absence of Wnt ligands, β-catenin is targeted for proteasomal degradation. However, upon Wnt activation, β-catenin is dissociated from the destructive complex and translocates to the nucleus.Adapted with permission from [132] © Elsevier (2012).
(Enlarge Image)
Figure 2.
Interaction map of Wnt pathway alteration in germ-free versus colonized mice. Wnt5a, Fzd3, Fzd8, sFrp2, sFrp4, Dkk3 and Wif1 are upregulated while Wnt8b is downregulated in the colon of germ-free mice. Green circles represent the significantly altered genes (p < 0.05).Data taken from [201].
Microbiota & Intestinal Stem Cells
The human intestinal tract is home to a complex ecosystem of commensal bacteria that live in a mutually beneficial state with the host. The number and composition of normal flora vary in different regions of the GI tract. A relatively low number and few species live in the stomach and upper small intestine due to the specific composition of luminal ingredients and the propulsive motion of the region. However, the distal part of the small intestine and colon is habitat to a diverse and densely populated microbiota. Phylogenetic analysis reveals a similar bacterial content in the distal ileum, ascending colon and rectum. The study of intestinal microbiota has gained considerable attention in recent years. The normal flora are composed of bacteria, archaea and fungi. To date, 500 different species of bacteria have been detected in the intestinal tract. The highest microbiota content is found in the distal ileum and colon where the bacterial concentration reaches 10–10 cells/ml of the luminal content. Under physiologic conditions, the host and microbiota retain a mutually beneficial symbiosis. However, any dysbiosis due to either microbiota composition alteration or host genetic susceptibility could initiate an unwarranted immune response.
The epithelial monolayer is covered with a mucus layer, which, together with the epithelium, forms the physical barrier against the luminal content. The mucus layer is 100–250 µm thick in the stomach and proximal small intestine, whereas it is 450 and 700 µm thick in the ileum and colon, respectively. The mucus layer is mainly composed of mucin glycoproteins and trefoil peptides. MUC2, MUC3, MUC12, MUC13 and MUC17 have been identified in the intestinal tract, with MUC2 being the predominant component of the mucus layer. While some mucins are secreted, some are membrane bound and reside on the apical surface of the epithelium. The mucus layer is composed of two distinct layers, both of which are mainly composed of MUC2. Mucus layers of the small intestine and colon are different in a number of ways. The small intestine does not have two distinct layers. In addition, the mucus layer is not firmly attached to the epithelium and can be relatively easily removed. By contrast, the mucus layer in the colon forms two distinct layers; the inner of which is firmly attached to the epithelium. The outer layer is more voluminous and is loosely attached to the inner layer. Under normal conditions, the inner layer is impermeable to bacteria.
In the noninflamed colon mucosa, bacteria are normally present within the crypts in the proximal colon. Moreover, the microbial composition of the crypts is found to be distinct from that of the lumen. However, the effect of this crypt-specific core microbiota on tissue homeostasis is currently unknown. In addition, despite knowledge on the fecal microbiota composition alteration in CRC and IBD patients compared with healthy controls, it is not known whether the crypt-specific core microbiota is also altered. However, a significant difference in microbiota composition has been observed between stool and mucosal biopsy samples in IBD. Spatial organization of bacteria is found to be altered in dextran sulfate sodium (DSS)-induced and IL-10 spontaneous colitis in mice and in IBD patients, as a consequence of which higher numbers of bacteria are found in the crypts. This change brings the bacteria into the close vicinity of stem cells at the base of crypts. Occasionally, intracellular bacteria are also observed in single epithelial cells; the pathologic relevance of this to intestinal homeostasis is unknown.
A number of studies have investigated the effect of intestinal injury by nonlethal intestinal bacteria on gut homeostasis and ISC in Drosophila. A net increase in the quantity of undifferentiated cells (ISCs and enteroblasts) is observed after intestinal injury. However, the number of ISCs is only mildly increased. In other words, the injury results in augmented stem cell division without significant expansion of the stem cell compartment. In addition, the stem cell features – that is, self-renewal and differentiation capacity – are not affected. Interestingly, TLR signaling is not altered at the transcriptomic level in this study. Buchon et al. suggest that ISCs are responsive to the damage signals rather than bacteria. For example, the effect of nonlethal intestinal infection on stem cell division is found to be mediated by reactive oxygen species and unpaired3, a component of the JAK/STAT pathway, which is produced by the damaged cells. The direct effect of microbiota on ISCs has been studied in axenic Drosophila. Axenic flies demonstrate a lower level of stem cell division and epithelium turnover, which are suggestive of the effect of microbiota on the ISC physiology. Notably, unpaired3 is completely absent in axenic flies while it is expressed in conventional counterparts.
Despite the recently elucidated significant difference in the genomic response to systemic inflammation in humans and mice, murine models are widely used to mimic human conditions. When wild-type mice are colonized with Salmonella typhimurium, intestinal epithelial proliferation is increased, which is more predominant at the base of the crypts. In addition, the intestinal expression of Wnt and Notch pathway components are affected.S. typhimurium has been shown to activate the Wnt signaling pathway both in vitro and in vivo. Short-term in vitro incubation of rat small intestinal epithelial IEC-18 with S. typhimurium results in a one- to sixfold increase in the gene expression level of Wnt2, Wnt2b, Wnt3a, Wnt7b and Wnt10b. The same experiment in HCT116 cells results in a two- to tenfold increase in the gene expression level of Fzd3, Fzd4, Fzd6, Fzd9 and Fzd10. Wnt2 and Fzd9 showed the highest alteration at the gene expression level. Interestingly, when human epithelial cells were manipulated to overexpress Wnt2, a lower level of NF-κB activity and downstream cytokine secretion was observed after S. typhimurium incubation. Intestinal Wnt2 expression at the protein level was also found to be increased in germfree mice colonized with S. typhimurium compared with control. Axin1 is a negative regulator of Wnt pathway. In vitro incubation of HCT116, HT29C19A and CaCo2BBE human intestinal cell lines with S. typhimurium significantly decreased Axin1 at the protein level while the gene expression level remained unchanged. This alteration was specific to S. typhimurium and was not reproduced by a proinflammatory stimulus such as TNF-α or nonpathogenic bacteria including Escherichia coli F18 and Lactobacillus rhamnosus GG. Similar reduction in Axin1 protein level was also observed in vivo. The effect of S. typhimurium on the intestinal epithelium was found to be via post-transcriptional modification of Axin1. Larsson et al. have conducted systematic intestinal gene expression analysis of germfree versus colonized mice. We queried the publicly available data for Wnt pathway ligands and receptors (Figure 1>). The genes that were significantly altered (p < 0.05) in the colon of germfree versus colonized mice were noted. Wif1, Wnt5a, Wnt8b, Fzd3, Fzd8, Sfrp4, Sfrp2 and Dkk3 were significantly altered in germfree mice compared with conventional counterparts. The protein–protein interaction map of these proteins with other members of the Wnt pathway is illustrated in Figure 2. The net effect of microbiota on the Wnt pathway cannot be inferred from the available data. Nonetheless, the data clearly show a potential regulatory mechanism of the Wnt pathway by the microbiota and plausibly by any alteration in the microbiota composition.
(Enlarge Image)
Figure 1.
Canonical Wnt signaling pathway. Wnt ligands undergo lipid modification in the endoplasmic reticulum of the Wnt-producing cells. PORCN is expressed in Wnt-secreting cells and is required for Wnt maturation and secretion. On the target cell, Wnt ligand binds to the Frizzled/LRP5,6 heterodimer. GSK and CK1 are kinases that phosphorylate components of the Wnt signaling cascade. sFRPs and WIF bind to Wnt ligands and DKK binds to LRP5/6, and thus inhibits pathway activation. Norrin and R-spondin are not components of the canonical Wnt pathway. However, they function via FZD/LRP and augment Wnt signaling. LGR is a R-spondin receptor and physically interacts with the FZD/LRP complex. In the absence of Wnt ligands, β-catenin is targeted for proteasomal degradation. However, upon Wnt activation, β-catenin is dissociated from the destructive complex and translocates to the nucleus.Adapted with permission from [132] © Elsevier (2012).
(Enlarge Image)
Figure 2.
Interaction map of Wnt pathway alteration in germ-free versus colonized mice. Wnt5a, Fzd3, Fzd8, sFrp2, sFrp4, Dkk3 and Wif1 are upregulated while Wnt8b is downregulated in the colon of germ-free mice. Green circles represent the significantly altered genes (p < 0.05).Data taken from [201].